CN113006188B - Excavator staged power matching method based on LSTM neural network - Google Patents
Excavator staged power matching method based on LSTM neural network Download PDFInfo
- Publication number
- CN113006188B CN113006188B CN202110301742.8A CN202110301742A CN113006188B CN 113006188 B CN113006188 B CN 113006188B CN 202110301742 A CN202110301742 A CN 202110301742A CN 113006188 B CN113006188 B CN 113006188B
- Authority
- CN
- China
- Prior art keywords
- excavator
- neural network
- stage
- lstm neural
- engine
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- E—FIXED CONSTRUCTIONS
- E02—HYDRAULIC ENGINEERING; FOUNDATIONS; SOIL SHIFTING
- E02F—DREDGING; SOIL-SHIFTING
- E02F9/00—Component parts of dredgers or soil-shifting machines, not restricted to one of the kinds covered by groups E02F3/00 - E02F7/00
- E02F9/20—Drives; Control devices
- E02F9/2025—Particular purposes of control systems not otherwise provided for
-
- E—FIXED CONSTRUCTIONS
- E02—HYDRAULIC ENGINEERING; FOUNDATIONS; SOIL SHIFTING
- E02F—DREDGING; SOIL-SHIFTING
- E02F9/00—Component parts of dredgers or soil-shifting machines, not restricted to one of the kinds covered by groups E02F3/00 - E02F7/00
- E02F9/20—Drives; Control devices
- E02F9/22—Hydraulic or pneumatic drives
- E02F9/2246—Control of prime movers, e.g. depending on the hydraulic load of work tools
-
- E—FIXED CONSTRUCTIONS
- E02—HYDRAULIC ENGINEERING; FOUNDATIONS; SOIL SHIFTING
- E02F—DREDGING; SOIL-SHIFTING
- E02F9/00—Component parts of dredgers or soil-shifting machines, not restricted to one of the kinds covered by groups E02F3/00 - E02F7/00
- E02F9/20—Drives; Control devices
- E02F9/22—Hydraulic or pneumatic drives
- E02F9/2264—Arrangements or adaptations of elements for hydraulic drives
-
- E—FIXED CONSTRUCTIONS
- E02—HYDRAULIC ENGINEERING; FOUNDATIONS; SOIL SHIFTING
- E02F—DREDGING; SOIL-SHIFTING
- E02F9/00—Component parts of dredgers or soil-shifting machines, not restricted to one of the kinds covered by groups E02F3/00 - E02F7/00
- E02F9/20—Drives; Control devices
- E02F9/22—Hydraulic or pneumatic drives
- E02F9/2264—Arrangements or adaptations of elements for hydraulic drives
- E02F9/2267—Valves or distributors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Structural Engineering (AREA)
- Civil Engineering (AREA)
- Mining & Mineral Resources (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Operation Control Of Excavators (AREA)
Abstract
The invention discloses an LSTM neural network-based excavator staged power matching method, and belongs to the technical field of engineering machinery. The realization process is as follows: the method comprises the steps of firstly collecting pressure data of a main pump, extracting characteristic vectors of two main pumps after filtering and denoising, mutually corresponding to each working stage to construct a sample space, inputting an established LSTM neural network model, adjusting learning rate and forgetting coefficient sampling interval parameters until accuracy meets requirements, inputting the model into an excavator control unit, monitoring the current stage of the excavator in real time, adjusting output torque and rotating speed of an engine and discharge capacity of the pump by using a second-order fuzzy regulator, ensuring that the engine is always in an economic working area, and finally realizing energy-saving control of the excavator.
Description
Technical Field
The invention relates to the technical field of engineering machinery, in particular to a method for automatically matching power according to the working stage of a hydraulic excavator.
Background
The excavator is widely applied to the industries of machinery, construction, transportation, water conservancy infrastructure military industry and the like. The existing excavator usually adopts step power control, and a driver can be divided into light load, economic or heavy load working modes according to different working objects. The excavator can be roughly divided into five different working stages of excavation, lifting and rotating, unloading, empty bucket returning and excavation preparation due to complex working conditions during operation. The power required for each operating phase is different. However, the operator often selects a larger gear to match all working stages, so that the output power of the engine is larger than the absorption power of the pump, the power of the pump of the engine is not matched, and the oil consumption of the excavator is higher. Therefore, if different working stages can be dynamically identified, the output rotating speed and the torque of the engine can be adjusted at any time, so that the engine and the hydraulic pump can be better matched in power, the operating performance of the excavator is improved, and the fuel consumption is reduced.
The main methods for identifying the working state of the excavator at the present stage comprise the following steps: the other method is that a sensor is used for detecting a pilot control signal of the excavator, the average value of pressure in a period and the percentage of the pressure larger than the average value are calculated, and the excavator is judged to be in a stone working condition, an earth working condition or a crushing working condition at present according to the percentage and the pressure signal of a rodless cavity of a bucket cylinder. The method can only roughly judge three working conditions and is lack of subdivision for each working condition. The second method is to shoot the current working process by using a camera installed on the excavator, match the picture with a database model and calculate which working stage the picture is in currently. The method additionally adds a camera, improves the cost of the whole machine, and cannot obtain effective load information. Besides, an SVM (support vector machine) is utilized to detect which working stage the pressure of the outlet of the excavator is currently in a classified mode, but an SVM neural network only uses the information of the current stage, and the model does not pay attention to the information at the last moment or even earlier, so that the accurate recognition rate of the model is low.
The excavator used by the invention adopts a double-pump oil supply system, the bucket rod oil cylinder, the rotary motor and the left walking motor are supplied with oil by the front pump, and the bucket oil cylinder, the movable arm oil cylinder and the right walking motor are supplied with oil by the rear pump. Some working stages adopt double pumps to supply oil simultaneously so as to improve the working speed, such as a lifting stage and a rotation stage.
Disclosure of Invention
The invention aims to provide a staged power matching method for an excavator, which is characterized in that historical data and current data of outlet pressure of the excavator are combined by an LSTM neural network to judge which working stage the excavator is currently in, and further carry out power matching control on an engine pump.
The purpose of the invention is realized by the following technical scheme:
an excavator staged power matching method based on an LSTM neural network is divided into two aspects: an excavator working stage identification method and a power matching method;
the excavator working stage identification method comprises the following steps:
s1: collecting original data, carrying out denoising processing on the original data, and extracting a pressure data signal;
s2: preprocessing the extracted pressure data signal to construct an input characteristic vector;
s3: corresponding data needing to be input into the LSTM neural network model to each stage of the excavator, enabling the data to be provided with labels related to each stage, and constructing a sample space;
s4: dividing the data with the labels into a training set, a cross validation set and a test set;
s5: designing an LSTM neural network model, and inputting label data into the neural network model for training;
s6: continuously adjusting learning rate sampling interval parameters on the cross validation set, and selecting parameter values corresponding to the model with the highest accuracy;
s7: and determining the final parameter value of the LSTM neural network model, and operating the LSTM neural network model on the test set data, wherein the operation result is the final excavator working stage identification accuracy.
The specific process of the power matching method is as follows: and identifying which working stage the excavator is in at present by the LSTM neural network model, and matching the power of the engine pump valve according to preset power in the corresponding current stage.
The collecting of the original data in the step S1 includes collecting pump outlet pressures of the main pump 1 and the main pump 2 of the excavator, where the sampling frequency is 100Hz, and after filtering and denoising, sliding and extracting a signal of a current window by using a time window with a width of 1S at a certain time interval.
The preprocessing of step S2 includes calculating the mean and variance of the data(ii) a Constructing the input feature vector includes: calculating the average value x of the pressure of the main pump 11(ii) a Calculating the average value x of the pressure of the main pump 22(ii) a Calculate the mean square error x of the main pump 1 pressure3(ii) a Calculating the mean square error x of the pressure of the main pump 24(ii) a Calculating the difference x between the average pressure values of the main pump 1 and the main pump 25(ii) a Calculating the difference x between the mean square values of the pressures of the main pump 1 and the main pump 26(ii) a Constructing an input feature vector as follows: x is the number oft=[x1,x2,x3,x4,x5,x6]。
The step S3 data label includes five stages of excavator working process: a digging preparation stage, a digging stage, a lifting rotation stage, an unloading stage and an empty bucket returning stage; the labels of the constructed input LSTM neural network model are represented using the following vectors: a digging preparation stage [1,0,0,0,0 ]; excavation stage [0,1,0,0,0 ]; lifting the slewing phase [0,0,1,0,0 ]; unload phase [0,0,0,1,0 ]; the empty bucket returns to stage [0,0,0,0,1 ].
The LSTM neural network structure adopts the following structure:
forget the door: f. oft=σ(Wf[ht-1,xt]+bf) Wherein W isfRepresents the forgetting gate weight coefficient, ht-1Representing the output of the hidden layer at the previous moment, bfRepresenting forgetting bias, sigma representing sigmoid activation function
An input gate: i.e. it=σ(Wi[ht-1,xt]+bi) Wherein W isiRepresenting the input gate weight coefficient, biRepresenting a deviation input to the hidden layer;
outputting memory information: ct=it*tanh(Wc[ht-1,xt]+bc)+ft*Ct-1Wherein W iscRepresenting the weight of the memory cell, bcRepresenting the deviation of the input layer to the memory cell;
an output gate: o ═ σ (W)o[ht-1,xt]+bo) Wherein W isoRepresenting the output gate weight coefficient, boRepresenting the deviation of the hidden layer from the output gate, the hidden layer output at time t being ht=Ot*tanh(Ct);
The neural network model is trained using the Adam optimization algorithm to update the network weights and biases and the time interval of sampling according to the gradient of the loss function.
Further, the specific process of the power matching method is as follows: once the LSTM neural network model identifies that the excavator is in an excavating stage, the working point of the engine is immediately adjusted to increase the output rotating speed and torque of the engine, once the LSTM neural network model identifies that the excavator is in a lifting and rotating stage, the output torque of the engine is unchanged, the displacement of a hydraulic pump is increased, the opening of a valve core is increased, when the LSTM neural network model identifies that the excavator is in an unloading stage, the rotating speed and torque of the engine are reduced, when the LSTM neural network model identifies that the excavator is in an empty bucket returning stage, the working point of the engine is further adjusted to reduce the displacement of the hydraulic pump, the opening of the valve core is reduced, and when the LSTM neural network model identifies that the excavator is in an excavating preparation stage, the output torque and the rotating speed of the engine and the displacement and the opening of the valve core of the hydraulic pump are further reduced.
The engine working point switching adopts a second-order fuzzy control algorithm, and the actual rotating speed of the engine, the rotating speed difference set in each working stage and the change rate of the rotating speed difference are taken as the input of a rotating speed regulator; taking the difference between the actual torque of the engine and the set torque and the change rate of the torque difference as the input of a torque regulator; the displacement of the hydraulic pump and the opening of the valve port are controlled by current signals.
Compared with the prior art, the invention has the following beneficial effects:
the invention can optimize the power of the engine pump at each stage in the excavating process of the excavator, makes up the defect that the power can only be matched under different working conditions in the prior art, and further reduces the energy consumption of the excavator. Compared with the traditional SVM neural network, the LSTM neural network improves the accurate recognition rate.
Drawings
FIG. 1 is a flow chart for excavator staged power matching
FIG. 2 is a diagram of the LSTM neural network model architecture
Detailed Description
The invention relates to an excavator working phase identification method and real-time power matching, wherein the excavator is specifically of the type XE200DA creep work, an engine selects QSB7-C227(S030077) of the Cornst corporation, a plunger type hydraulic pump further comprises two pressure sensors and a controller, and the two pressure sensors are respectively arranged on two main pumps.
The method comprises the following steps:
1 data sampling and filtering process
Two main pump outlet pressure waveform data are collected from the beginning of a digging preparation stage, the sampling frequency is 100Hz, and noise in signals is removed by adopting a weighted average filtering method. y is0Is the sample value at the current time t, yiIs the sampled value at time deltat of t-i,for the filtered values then the formula can be expressed as:
in the formula CiThe weight of the coefficient representing different time points is larger as the coefficient approaches the current time point, and the following conditions are satisfied
The pressure signals of the two main pumps are extracted using a time window sliding of width 1 s.
2 constructing neural network input and output vectors
The mean value of the main pump 1, the main pump 2 and the mean square error of the main pump 1 and the main pump 2, and the mean value of the difference value of the main pump 1 and the mean square error of the difference value of the main pump 2 are calculated as input vectors, respectively. The formula is as follows:
forming six calculation results into input vector x of neural networkt=[x1,x2,x3,x4,x5,x6];
Normalizing the input vector: wherein x ismax=max(x1,x2,x3,x4,x5,x6),xmin=(x1,x2,x3,x4,x5,x6);
The output vector corresponds to each stage of excavator work: a digging preparation stage [1,0,0,0,0 ]; excavation stage [0,1,0,0,0 ]; lifting the slewing phase [0,0,1,0,0 ]; unload phase [0,0,0,1,0 ]; the empty bucket returns to stage [1,0,0,0,0 ].
3 constructing a neural network model
The LSTM neural network structure meets the following requirements:
forget the door: f. oft=σ(Wf[ht-1,xt]+bf) Wherein W isfRepresents the forgetting gate weight coefficient, ht-1Representing the output of the hidden layer at the previous moment, bfRepresenting forgetting bias, sigma representing sigmoid activation function
An input gate: i.e. it=σ(Wi[ht-1,xt]+bi) Wherein W isiRepresenting the input gate weight coefficient, biRepresenting a deviation input to the hidden layer;
outputting memory information: ct=it*tanh(Wc[ht-1,xt]+bc)+ft*Ct-1Wherein W iscRepresenting the weight of the memory cell, bcRepresenting the deviation of the input layer to the memory cell;
an output gate: o ═ σ (W)o[ht-1,xt]+bo) Wherein W isoRepresenting the output gate weight coefficient, boRepresenting the deviation of the hidden layer from the output gate, the hidden layer output at time t being ht=Ot*tanh(Ct);
The neural network model is trained to update network weights and biases and sampling time intervals according to the gradient of the loss function by using an Adam optimization algorithm.
Collecting pressure waveforms of 200 main pumps, constructing 1000 characteristic vectors, forming a sample space together with corresponding output, dividing 500 into training sets, 300 into cross validation sets, and 200 into test sets, training on a PC, and inputting the trained network model into a control unit.
4 excavator staged power matching
Once the LSTM neural network model identifies that the excavator is in an excavating stage, the working point of the engine is immediately adjusted to increase the output rotating speed and torque of the engine, once the LSTM neural network model identifies that the excavator is in a lifting and rotating stage, the output torque of the engine is unchanged, the displacement of a hydraulic pump is increased, the opening of a valve core is increased, when the LSTM neural network model identifies that the excavator is in an unloading stage, the rotating speed and torque of the engine are reduced, when the LSTM neural network model identifies that the excavator is in an empty bucket returning stage, the working point of the engine is further adjusted to reduce the displacement of the hydraulic pump, the opening of the valve core is reduced, and when the LSTM neural network model identifies that the excavator is in an excavating preparation stage, the output torque and the rotating speed of the engine and the displacement and the opening of the valve core of the hydraulic pump are further reduced.
The engine working point switching adopts a second-order fuzzy control algorithm, and the actual rotating speed of the engine, the rotating speed difference set in each working stage and the change rate of the rotating speed difference are taken as the input of a rotating speed regulator; taking the difference between the actual torque of the engine and the set torque and the change rate of the torque difference as the input of a torque regulator; the displacement of the hydraulic pump and the opening of the valve port are controlled by current signals.
Claims (7)
1. An excavator staged power matching method based on an LSTM neural network is characterized by comprising the following two aspects: an excavator working stage identification method and a power matching method;
the excavator working phase identification method comprises the following steps:
s1: collecting original data, carrying out denoising processing on the original data, and extracting a pressure data signal;
s2: preprocessing the extracted pressure data signal to construct an input characteristic vector;
s3: corresponding data needing to be input into the LSTM neural network model to each stage of the excavator, enabling the data to be provided with labels related to each stage, and constructing a sample space;
s4: dividing the data with the labels into a training set, a cross validation set and a test set;
s5: designing an LSTM neural network model, and inputting label data into the LSTM neural network model for training;
s6: continuously adjusting learning rate and sampling interval parameters on the cross validation set, and selecting a parameter value corresponding to the model with the highest accuracy;
s7: determining a final parameter value of the LSTM neural network model, and operating the LSTM neural network model on the test set data, wherein an operation result is the final excavator working stage identification accuracy;
the specific process of the power matching method is as follows: and identifying which working stage the excavator is in at present by the LSTM neural network model, and matching the power of the engine pump valve according to preset power in the corresponding current stage.
2. The LSTM neural network based excavator staged power matching method of claim 1, wherein: the collecting of the raw data in the step S1 includes collecting pump outlet pressures of the main pump 1 and the main pump 2 of the excavator, where the sampling frequency is 100Hz, and after filtering and denoising, the signals are extracted by sliding with a window of a certain width at a certain interval.
3. The LSTM neural network based excavator staged power matching method of claim 1, wherein: the preprocessing of step S2 includes calculating the mean and variance of the data; the constructing of the input feature vector comprises: calculating the average value x of the pressure of the main pump 11(ii) a Calculating the average value x of the pressure of the main pump 22(ii) a Calculate the mean square error x of the main pump 1 pressure3(ii) a Calculating the mean square error x of the pressure of the main pump 24(ii) a Calculating the average value x of the pressure difference between the main pump 1 and the main pump 25(ii) a Calculating the mean square value x of the pressure difference between the main pump 1 and the main pump 26(ii) a Constructing an input feature vector as follows: x is the number oft=[x1,x2,x3,x4,x5,x6]。
4. The LSTM neural network based excavator staged power matching method of claim 1, wherein: the step S3 data label includes five stages of excavator working process: a digging preparation stage, a digging stage, a lifting rotation stage, an unloading stage and an empty bucket returning stage; the labels of the constructed input LSTM neural network model are represented using the following vectors: a digging preparation stage [1,0,0,0,0 ]; excavation stage [0,1,0,0,0 ]; lifting the slewing phase [0,0,1,0,0 ]; unload phase [0,0,0,1,0 ]; the empty bucket returns to stage [0,0,0,0,1 ].
5. The LSTM neural network based excavator staged power matching method of claim 1, wherein: the LSTM network comprises the following structure:
forget the door: f. oft=σ(Wf[ht-1,xt]+bf) Wherein W isfRepresents the forgetting gate weight coefficient, ht-1Representing the output of the hidden layer at the previous moment, bfRepresenting forgetting bias, sigma representing sigmoid activation functionAn input gate: i.e. it=σ(Wi[ht-1,xt]+bi) Wherein W isiRepresenting the input gate weight coefficient, biRepresenting a deviation input to the hidden layer;
outputting memory information: ct=it*tanh(Wc[ht-1,xt]+bc)+ft*Ct-1Wherein W iscRepresenting the weight of the memory cell, bcRepresenting the deviation of the input layer to the memory cell;
an output gate: o ═ σ (W)o[ht-1,xt]+bo) Wherein W isoRepresenting the output gate weight coefficient, boRepresenting the deviation of the hidden layer from the output gate, the hidden layer output at time t being ht=Ot*tanh(Ct);
The neural network model is trained using the Adam optimization algorithm to update the network weights and biases and the time interval of sampling according to the gradient of the loss function.
6. The grading power matching method for the excavator based on the LSTM neural network as claimed in claim 1, wherein the specific process of the power matching method is as follows: once the LSTM neural network model identifies that the excavator is in the excavation stage, immediately adjusting the working point of the engine, and increasing the output rotating speed and torque of the engine; once the LSTM neural network model identifies that the excavator is in a lifting rotation stage, the output torque of the engine is unchanged, the displacement of the hydraulic pump is increased, and the opening of the valve core is increased; when the LSTM neural network model identifies that the excavator is in an unloading stage, reducing the rotating speed and torque of the engine; when the LSTM neural network model identifies that the excavator is in an empty bucket return stage, further reducing the working point of the engine, reducing the displacement of the hydraulic pump and reducing the opening of the valve core; when the LSTM neural network model identifies that the excavator is in the excavation preparation stage, the output torque and the rotating speed of the engine and the displacement of the hydraulic pump are further reduced.
7. The grading power matching method of the excavator based on the LSTM neural network as claimed in claim 6, wherein the engine operating point switching adopts a second order fuzzy control algorithm, and the actual rotating speed of the engine, the rotating speed difference set in each operating stage and the change rate of the rotating speed difference are taken as the input of the rotating speed regulator; taking the difference between the actual torque of the engine and the set torque and the change rate of the torque difference as the input of a torque regulator; the displacement of the hydraulic pump and the opening of the valve core are controlled by current signals.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110301742.8A CN113006188B (en) | 2021-03-22 | 2021-03-22 | Excavator staged power matching method based on LSTM neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110301742.8A CN113006188B (en) | 2021-03-22 | 2021-03-22 | Excavator staged power matching method based on LSTM neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113006188A CN113006188A (en) | 2021-06-22 |
CN113006188B true CN113006188B (en) | 2022-02-18 |
Family
ID=76404137
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110301742.8A Active CN113006188B (en) | 2021-03-22 | 2021-03-22 | Excavator staged power matching method based on LSTM neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113006188B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE102023200194A1 (en) | 2023-01-11 | 2024-07-11 | Hawe Hydraulik Se | Method for automated, measurement data-based design of an electronic controller for a hydraulic system |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113502870B (en) * | 2021-07-15 | 2023-02-07 | 上海三一重机股份有限公司 | Excavator working condition judging method and device |
CN113591216B (en) * | 2021-07-23 | 2023-11-17 | 三一重机有限公司 | Excavator working mode determining method and system and excavator |
CN113640024B (en) * | 2021-08-09 | 2024-08-20 | 上海华兴数字科技有限公司 | Excavator fault diagnosis method, excavator, electronic device and storage medium |
CN113774985B (en) * | 2021-10-28 | 2023-03-24 | 三一重机有限公司 | Method and device for controlling working machine, medium, and working machine |
CN114407680B (en) * | 2021-12-31 | 2024-03-01 | 柳州柳工挖掘机有限公司 | Estimation method for remaining driving mileage of electric excavator |
CN114357880B (en) * | 2022-01-04 | 2024-10-15 | 大连理工大学 | Staged hydraulic excavator engine working point optimizing method |
CN115434819B (en) * | 2022-08-12 | 2024-04-12 | 中联重科股份有限公司 | Method and device for controlling power matching, electronic equipment and engineering machinery |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104074225A (en) * | 2014-07-08 | 2014-10-01 | 湖南机电职业技术学院 | Hydraulic excavator power adaptive control system and method |
KR101791233B1 (en) * | 2016-05-31 | 2017-10-31 | (재)건설기계부품연구원 | power matching system for excavator and controlling method of the same |
CN107688849A (en) * | 2017-07-28 | 2018-02-13 | 北京深鉴科技有限公司 | A kind of dynamic strategy fixed point training method and device |
CN108678065A (en) * | 2018-05-31 | 2018-10-19 | 马鞍山松鹤信息科技有限公司 | A kind of excavator energy-saving control method |
CN109636951A (en) * | 2018-11-21 | 2019-04-16 | 中南大学 | Excavator energy consumption analysis method based on working phase recognition |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018014718A1 (en) * | 2016-07-22 | 2018-01-25 | 厦门大学 | Method for recognizing difficulty level of working condition of loading machine |
-
2021
- 2021-03-22 CN CN202110301742.8A patent/CN113006188B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104074225A (en) * | 2014-07-08 | 2014-10-01 | 湖南机电职业技术学院 | Hydraulic excavator power adaptive control system and method |
KR101791233B1 (en) * | 2016-05-31 | 2017-10-31 | (재)건설기계부품연구원 | power matching system for excavator and controlling method of the same |
CN107688849A (en) * | 2017-07-28 | 2018-02-13 | 北京深鉴科技有限公司 | A kind of dynamic strategy fixed point training method and device |
CN108678065A (en) * | 2018-05-31 | 2018-10-19 | 马鞍山松鹤信息科技有限公司 | A kind of excavator energy-saving control method |
CN109636951A (en) * | 2018-11-21 | 2019-04-16 | 中南大学 | Excavator energy consumption analysis method based on working phase recognition |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE102023200194A1 (en) | 2023-01-11 | 2024-07-11 | Hawe Hydraulik Se | Method for automated, measurement data-based design of an electronic controller for a hydraulic system |
Also Published As
Publication number | Publication date |
---|---|
CN113006188A (en) | 2021-06-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113006188B (en) | Excavator staged power matching method based on LSTM neural network | |
CN109636951B (en) | Excavator energy consumption analysis method based on working phase recognition | |
CN102877500B (en) | Automatic identification method for operating cycle stages of excavator | |
WO2006016440A1 (en) | Data processing method, data processing device, diagnosis method, and diagnosis device | |
CN112096696B (en) | Self-adaptive inversion control method for pump-controlled asymmetric hydraulic position system | |
CN103177155A (en) | Oilfield pumping unit oil pumping energy saving and production increasing optimization method based on back propagation neural network (BPNN) and strength Pareto evolutionary algorithm 2 (SPEA2) | |
CN113374021B (en) | Excavator working condition identification method based on pilot control signal of operating handle | |
CN110826790A (en) | Intelligent prediction method for construction productivity of cutter suction dredger | |
CN113266318B (en) | Self-learning-based stroke frequency adjusting method for rod-type pumping unit well | |
Jin et al. | Energy saving strategy of the variable-speed variable-displacement pump unit based on neural network | |
CN112925208A (en) | Disturbance compensation method for data-driven electro-hydraulic servo system of well drilling machine | |
CN114357880B (en) | Staged hydraulic excavator engine working point optimizing method | |
CN111648758B (en) | Model-free self-adaptive control method and system for well drilling machine propulsion device | |
Shi et al. | Working stage identification of excavators based on control signals of operating handles | |
CN114792067B (en) | Method for classifying and identifying working phases of excavator based on ResNet neural network | |
CN114091339A (en) | Method for predicting bow blowing instantaneous yield of drag suction dredger based on GRU | |
CN115012469A (en) | Intelligent control system and method for hydraulic excavator without gear control and with self-adaptive working condition | |
CN116976146B (en) | Fracturing well yield prediction method and system coupled with physical driving and data driving | |
CN109610549B (en) | Frequency conversion intelligent control system and control method for hydraulic excavator electric appliance | |
CN108804721B (en) | Oil pumping machine fault diagnosis method based on self-adaptive unscented Kalman filtering and RBF neural network | |
CN114862058B (en) | Optimal control method and system for pipeline conveying system of dredger | |
CN113431925A (en) | Electro-hydraulic proportional valve and position control system, control method and fault prediction method thereof | |
CN114351785B (en) | Hydraulic excavator system flow matching optimization method based on reinforcement learning | |
He et al. | An approach to recognizing the working cycle stage with small sample data for energy conservation | |
CN113844861B (en) | Automatic speed regulating method of belt and coal mining machine based on coal quantity detection |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |